Physics Fluctuomatics (Tohoku University) 1 Physical Fluctuomatics 7th~10th Belief propagation...

43
Physics Fluctuomatics (Tohoku University) 1 Physical Fluctuomatics 7th~10th Belief propagation Appendix Kazuyuki Tanaka Graduate School of Information Sciences, Tohoku University [email protected] http://www.smapip.is.tohoku.ac.jp/~kazu/

Transcript of Physics Fluctuomatics (Tohoku University) 1 Physical Fluctuomatics 7th~10th Belief propagation...

Physics Fluctuomatics (Tohoku University) 1

Physical Fluctuomatics7th~10th Belief propagation

Appendix

Kazuyuki TanakaGraduate School of Information Sciences, Tohoku University

[email protected]://www.smapip.is.tohoku.ac.jp/~kazu/

Physics Fluctuomatics (Tohoku University) 2

Textbooks

Kazuyuki Tanaka: Introduction of Image Processing by Probabilistic Models, Morikita Publishing Co., Ltd., 2006 (in Japanese) , Chapter 5.

ReferencesH. Nishimori: Statistical Physics of Spin Glasses and Information Processing, ---An Introduction, Oxford University Press, 2001. H. Nishimori, G. Ortiz: Elements of Phase Transitions and Critical Phenomena, Oxford University Press, 2011.M. Mezard, A. Montanari: Information, Physics, and Computation, Oxford University Press, 2010.

Physics Fluctuomatics (Tohoku University) 3

Probabilistic Model for Ferromagnetic MaterialsProbabilistic Model for

Ferromagnetic Materials

p p

p p

)1,1()1,1()1.1()1.1( PPPP

pPP )1.1()1,1(

11 a

1

12 a

1

11

1 1

p

PP

2

1

)1.1()1,1(

Physics Fluctuomatics (Tohoku University) 4

Probabilistic Model for Ferromagnetic MaterialsProbabilistic Model for

Ferromagnetic Materials

Prior probability prefers to the configuration with the least number of red lines.

> >=

Lines Red of #Lines Blue of # )2

1()( ppaP

p p

11 a 112 a 111 1 1

Physics Fluctuomatics (Tohoku University) 5

More is different in Probabilistic Model for Ferromagnetic Materials

Disordered State

Ordered State

Sampling by Markov Chain Monte Carlo method

p p

Small p Large p

p p

More is different.

p2

1p

2

1

Critical Point(Large fluctuation)

Physics Fluctuomatics (Tohoku University) 6

Fundamental Probabilistic Models for Magnetic Materials

Since h is positive, the probablity of up spin is larger than the one of down spin .

1

)exp(

)exp()(

a

ha

haaP

1a

+1 -1

he he

)tanh()(1

haaPma

h : External Field

)(tanh1)()( 2

1

2 haPmaaVa

Variance

Average

0h

Physics Fluctuomatics (Tohoku University) 7

Fundamental Probabilistic Models for Magnetic Materials

Since J is positive, (a1,a2)=(+1,+1) and (-1,-1) have the largest probability .

1 121

2121

1 2

)exp(

)exp(),(

a a

aJa

aJaaaP

11 a

0),(1 1

2111

1 2

a a

aaPam

J : Interaction

1),()(1 1

212

111

1 2

a a

aaPmaaVVariance

Average

0J

Je Je

+1 +1 -1 -1

+1 +1 -1-1

12 aJe Je

Physics Fluctuomatics (Tohoku University) 8

Fundamental Probabilistic Models for Magnetic Materials

a

aEZ

))(exp(

Eji

jiVi

i aaJahaE},{

)(

Translational Symmetry

),( EVJ

J

h h

)(exp1

)( aEZ

aP

),,,( 21 Naaaa

E : Set of All the neighbouring Pairs of Nodes

1ia 1ia

N

i ai aPa

Nm

1

)(1

Problem: Compute

)'()()'()( aPaPaEaE

Physics Fluctuomatics (Tohoku University) 9

Fundamental Probabilistic Models for Magnetic Materials

Eji

jiVi

i aaJahaE},{

)(

N

i ai

NhaPa

Nm

10)(

1limlim

)(exp1

)( aEZ

aP

),,,( 21 Naaaa

1ia

Problem: Compute

Translational Symmetry

),( EV

J

J

h h

Spontaneous Magnetization

Physics Fluctuomatics (Tohoku University) 10

Mean Field Approximation for Ising Model

)},{( 0))(( Ejimama ji We assume that the probability for configurations satisfying

Vi

iaJmhaE )4()(

2mmamaaa ijji

Eji

jiVi

i aaJahaE},{

iJm

Jm

JmJm

h

are large.

Physics Fluctuomatics (Tohoku University) 11

Mean Field Approximation for Ising Model

)4tanh()(1

1

JmhaPaN

mN

i ai

Vi

ii aPaEZ

aP )())(exp(1

)(

Fixed Point Equation of m)(mm

We assume that all random variables ai are independent of each other, approximately.

Vi

iaJmhaE )4()(

Physics Fluctuomatics (Tohoku University) 12

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM

Physics Fluctuomatics (Tohoku University) 13

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

0

xy

)(xy

y

x*M

Physics Fluctuomatics (Tohoku University) 14

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

0M0

xy

)(xy

y

x*M

Physics Fluctuomatics (Tohoku University) 15

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

01 MM

0M

1M

0

xy

)(xy

y

x*M

Physics Fluctuomatics (Tohoku University) 16

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

12

01

MM

MM

0M1M

1M

0

xy

)(xy

y

x*M

Physics Fluctuomatics (Tohoku University) 17

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

12

01

MM

MM

0M1M

1M

0

xy

)(xy

y

x*M

2M

Physics Fluctuomatics (Tohoku University) 18

Fixed Point Equation and Iterative Method

• Fixed Point Equation ** MM • Iterative Method

23

12

01

MM

MM

MM

0M1M

1M

0

xy

)(xy

y

x*M

2M

Physics Fluctuomatics (Tohoku University) 19

Marginal Probability Distribution in Mean Field Approximation

))4exp((1

)()(

1 2 1 1

ii

a a a a aii

aJmhZ

aPaP

i i N

i

JmJm

JmJm

h

1

)(

iaiii aPam

))4tanh(( mJhm Jm: Mean Field

Physics Fluctuomatics (Tohoku University) 20

Advanced Mean Field Method

))4exp((1

)( ii

ii ahZ

aP

)))(3exp((1

),( jijii

jiij aJaaahZ

aaP

l

h

ll

h

h

1

),()(

jajiijii aaPaP

))3tanh()(tanh(arctanh hJ

Bethe Approximation

Kikuchi Method (Cluster Variation Meth)

l : Effective Field

Fixed Point Equation for l

J

Physics Fluctuomatics (Tohoku University) 21

Average of Ising Model on Square Grid Graph

(a) Mean Field Approximation(b) Bethe Approximation(c) Kikuchi Method (Cluster Variation Method)(d) Exact Solution ( L. Onsager , C.N.Yang )

J/1

a

iNh

aPa

)(limlim

0

Ejiji

Vii aaJah

ZaP

},{

exp1 ),( EVJ

J

h h

Physics Fluctuomatics (Tohoku University) 22

Model Representation in Statistical Physics

),,,(},,,Pr{ 212211 NNN aaaPaAaAaA

a

aEZ

))(exp(

)(}Pr{ aPaA

))(exp(1

)( aEZ

aP

),,,( 21 NAAAA

Gibbs Distribution Partition Function

)))(exp(ln(ln a

aEZF

Free Energy

Energy Function

Physics Fluctuomatics (Tohoku University) 23

Gibbs Distribution and Free Energy

Gibbs Distribution

ZPFaQQFaQ

ln][}1)(|][{min

))(exp(1

)( aEZ

aP

)(ln)()()(][ aQaQaQaEQFaa

Variational Principle of Free Energy Functional F[Q] under Normalization Condition for Q(a)

Free Energy Functional of Trial Probability Distribution Q(a)

a

aEZ

))(exp(lnlnFree Energy

Physics Fluctuomatics (Tohoku University) 24

Explicit Derivation of Variantional Principle for Minimization of Free Energy Functional

ZPFaQQFaQ

ln][}1)(|][{min

)(

)(exp

)(exp)(ˆ aP

aE

aEaQ

a

1)()())(ln)((1)(

aaa

aQaQaQaEaQQFQL

01)(ln)(

)(

aQaEaQ

QL

1)(exp)(ˆ aEaQ

Normalization Condition

Physics Fluctuomatics (Tohoku University) 25

Kullback-Leibler Divergence and Free Energy

0)(

)(ln)(

aP

aQaQPQD

a

a

aQaQ

1)( ,0)(

ZQF

ZaQaQaEaQPQD

QF

aa

ln][

ln)(ln)()()(]|[

][

0)()( PQDaPaQ

))(exp(1

)( aEZ

aP

}1)(|]|[{minarg}1)(|][{minarg aQaQ

aQPQDaQQF

Physics Fluctuomatics (Tohoku University) 26

Interpretation of Mean Field Approximation as Information Theory

Vi

ii aQaQ )()(

)(

)(ln)(

aP

aQaQPQD

a

))(exp(1

aEZ

aP

and

Marginal Probability Distributions Qi(ai) are determined so as to minimize D[Q|P]

1 2 1 1 2

)()()(\ a a a a a aaa

ii

i i i Ni

aQaQaQ

Minimization of Kullback-Leibler Divergence between

Physics Fluctuomatics (Tohoku University) 27

Interpretation of Mean Field Approximation as Information Theory

Eji

jiVi

i aaJahaE},{

)(

Vi a

iiVi a

i

i

aPaV

aPaV

m1

)(||

1)(

||

1

)(exp1

)( aEZ

aP

),,,( ||21 Vaaaa

1ia

Problem: Compute

Translational Symmetry

),( EV

J

J

h h

Magnetization

1 2 1 1 2

)()()(\ a a a a a aaa

ii

i i i Ni

aPaPaP

Physics Fluctuomatics (Tohoku University) 28

Kullback-Leibler Divergence in Mean Field Approximation for Ising Model

Vi

ii aQaQ )()(

ZViQFPQD i ln|MF

)(

)(ln)(

aP

aQaQPQD

a

Viii

Ejiji

Viii

QQQQJ

QhViQF

1},{ 11

1MF

ln))()()((

)(}]|[{

1 2 1 1 2

)(

)()(\

a a a a a a

aaii

i i i N

i

aQ

aQaQ

Eji

jiVi

i aaJahaE},{

)( )(exp

1)( aEZ

aP

Physics Fluctuomatics (Tohoku University) 29

Minimization of Kullback-Leibler Divergence and Mean Field Equation

)( ))(ˆ(exp1ˆ

1

ViQJhZ

Qij

ji

i

} ,1)(|]|[{minarg)}(ˆ{}{

ViQPQDQ iQ

ii

Fixed Point Equations for {Qi|iV}

Variation

1 1

))(ˆ(exp

ij

ji QJhZ

i

Set of all the neighbouring nodes of the node i

Ejiji },{

Physics Fluctuomatics (Tohoku University) 30

Orthogonal Functional Representation of Marginal Probability Distribution of Ising Model

iiii amaQ2

1

2

1)(

1

)(

iaiii

aii aQaaQam

),,,( 21 Naaaa 1ia

ia

iiia

iia

iii

aii

ai

aii

iiii

maQadddacaaQa

aQccdacaQ

adacaQ

iii

iii

2

1)(

2

12)()(

2

1)(

2

12)()(

1)( )(

111

111

2

1 2 1 1

)()()(\ a a a a aaa

ii

i i Ni

aQaQaQ

Physics Fluctuomatics (Tohoku University) 31

Conventional Mean Field Equation in Ising Model

)4tanh( Jmhm

iiiiiaa

ii maamaQaPaPi

2

1

2

1

2

1

2

1)(ˆ)()(

\

maPaN

N

i ai

1

)(1

Fixed Point Equation

mmmm N 21

Eji

jiVi

i aaJahaE},{

)(

))4exp((1

))(ˆ(exp1

)(ˆ1

ii

iij

ji

ii aJmhZ

aQJhZ

aQ

VJ

J

Translational Symmetry

h h

)( 4|| Vii

Physics Fluctuomatics (Tohoku University) 32

Interpretation of Bethe Approximation (1)

Eji

jiVi

i aaJahaE},{

)(

)(exp1

)( aEZ

aP

),,,( ||21 Vaaaa

1ia

Translational Symmetry

),( EV

J

J

h h

1 2 1 1 2

)()()(\ a a a a a aaa

ii

i i i Ni

aPaPaP

1 2 1 1 2 1 1 2

)()(),(},\{ a a a a a a a a aaaa

jiij

i i i j j j Nji

aPaPaaP

Eji

jiij aaZ

aP},{

),(1

)(

jijijiij aJaha

jha

iaa

||

1

||

1exp),(

a Eji

jiij aaZ

},{

),(

Compute

and

Interpretation of Bethe Approximation (2)

ZQFPQD ln

aQaQaaWaaQ

aQaQaaWaQ

aQaQaaWaQQF

aEji a ajiijjiij

aEji a ajiij

aaa

aEjijiij

a

i j

i j ji

ln)(,ln,

ln)(,ln)(

ln)(,ln)(

},{

},{ ,\

},{

0ln)(

aP

aQaQPQD

a

Free EnergyKL Divergence

Eji

jiij aaWZ

P},{

,1

x

ji aaa

jiij

aQ

aaQ

,

)(

),(

\

33Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (3)

ZQFPQD ln

Ejijjiiijij

Viii

Ejiijij

a

Ejiijij

QQQQQQ

QQ

WQ

aQaQ

WQQF

},{

},{

},{

lnln,ln,

ln

,ln,

ln)(

,ln,

Bethe Free

Energy

Free EnergyKL Divergence

Eji

jiij aaWZ

aP},{

,1

ji aaajiij aQaaQ

,

)(),(\

iaaii aQaQ

\

)()(

34Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (4)

FPQDQQ

minargminarg

,iji QQ

ZQQFPQD iji ln,Bethe

ijiQQQ

QQFPQDiji

,minargminarg Bethe,

1,

iji QQ

Ejijjiiijij

Viii

Ejiijijiji

QQQQQQ

QQWQQQF

},{

},{Bethe

lnln,ln,

ln,ln,,

35Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (5)

Ejiijij

Viii

Vi ijijijii

ijiiji

QQ

QQ

QQFQQL

},{

},{,

BetheBethe

1,1

,

,,

1, ,,,minarg Bethe,

ijiijiijiQQ

QQQQQQFiji

Lagrange Multipliers to ensure the constraints

36Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (6)

Ejiijij

Viii

Vi ijijijji

Ejijjiiijij

Viii

Ejiijij

Ejiijij

Viii

Vi ijijijiiijiiji

QQQQ

QQQQQQ

QQQ

QQ

QQQQFQQL

},{},{,

},{

},{

},{

},{,BetheBethe

1,1,

lnln,ln,

ln,ln,

1,1

,,,

0,Bethe

iji

ii

QQLxQ

• Extremum Condition

0,, Bethe

ijijiij

QQLxxQ

37Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (7)

FGfP yxyxyx g ,,,

ik

ikiiii ai

aQ )(1||

1exp },{, )()(exp,, 2}2,1{,21}2,1{,121122112 aaaaaaQ

Extremum

Condition 0,Bethe

iji

ii

QQLxQ 0,

, Bethe

iji

jiij

QQLxxQ

38Physics Fluctuomatics (Tohoku

University)

115114

1131121

11

1

aMaM

aMaMZ

aQ

2282272262112

11511411312

2112

,

1,

aMaMaMaaW

aMaMaMZ

aaQ

)()(exp\

},{, ijik

ikijii aMa

Interpretation of Bethe Approximation (8)

FGfP yxyxyx g ,,,14 2

5

13M

14M

15M

12M

3

115114

1131121

111

aMaM

aMaMZ

aQ

2282272262112

11511411312

2112

,

1,

aMaMaMaaW

aMaMaMZ

aaQ

Extremum

Condition 0,Bethe

iji

ii

QQLxQ 0,

, Bethe

iji

jiij

QQLxxQ

39Physics Fluctuomatics (Tohoku

University)

26M

14

5

13M

14M

15M

12W3

2

6

27M

8

7

28M

Interpretation of Bethe Approximation (9)

FGfP yxyxyx g ,,,14 2

5

13M

14M

15M

12M

3

412W

1

5

13M

14M

15M

3

26M

2

6

27M

8

7

28M

,121 QQ

115114

1131121

11

1

aMaM

aMaMZ

aQ

2282272262112

11511411312

2112

,

1,

aMaMaMaa

aMaMaMZ

aaQ

1514

1312

21

,

MM

M

M

Message Update Rule

40Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (10)

15141312

15141312

21 ,

,

MMMW

MMMW

M

1

3

4 2

5

13M

14M

15M

21M

14

5

3

2

6

8

7

2a

14 2

5

3

=

Message Passing Rule of Belief Propagation

It corresponds to Bethe approximation in the statistical mechanics.

41Physics Fluctuomatics (Tohoku

University)

Interpretation of Bethe Approximation (11)

1 1 \

1 \

,

,

jikikij

jikikij

ji MW

MW

M

42Physics Fluctuomatics (Tohoku

University)

))tanh()(tanh(arctanh\

jik

ikji hJ

jijiM exp

))3tanh()(tanh(arctanh hJ

ji Translational Symmetry

Physics Fluctuomatics (Tohoku University) 43

Summary

Statistical Physics and Information TheoryProbabilistic Model of FerromagnetismMean Field TheoryGibbs Distribution and Free EnergyFree Energy and Kullback-Leibler DivergenceInterpretation of Mean Field Approximation as Information TheoryInterpretation of Bethe Approximation as Information Theory